We have simplified the dataset and made it ready for extension by deprecating the IEEEXplore article numbers. Instead, the DOI column should now be used as a unique identifier. Where a paper did not include a valid DOI (2 cases at this point), a syntactically valid but fake DOI was entered that starts with 10.0000/The reference column has now also been updated to point to DOIs instead of the article numbers. At this point you can still access the article number columns but they have been moved to the last columns of the table.
We added a new column R which contains the author keywords supplied for each paper. The author keywords are words chosen as representative for the work by the authors of the paper. They can be found on the paper pdf. Where no keyword is present the authors did not provide these keywords.
We updated the dataset with the 2015 IEEE VIS
If you are interested in studying the history of VAST you're pretty safe just using the conference name column. If you are interested in studying the history of InfoVis - you're more or less safe depending on what you're interested in. If you're interested in a completeness of papers and go by the conference name column alone, you may miss a few papers that are among the VIS papers (maybe a handful a year). If you would like to track them down you could go into the proceedings and find the respective InfoVis sessions and mark those papers as InfoVis-related in the dataset (feel free to send us your data updates for inclusion here). If you're interested in the evolution of SciVis you now have to use a combination of Vis and SciVis conference names but beware that you will be including a few papers that nowadays people would put into InfoVis or VAST category.
In order to give a more correct reflection of the historical evolution of the field we decided to introduce "Vis" back into the conference names column for all papers that met the following criteria: conference name = SciVis, year <= 2012. This better reflects the following:What does this mean for you?
There was only one conference in the beginning. This conference was called Vis. It included a wide variety of papers, including what some would call InfoVis or VAST today. Even with the foundation of the InfoVis symposium, in the beginning it was considered more prestigious to send your best work to the "main" conference which was the Vis conference. This culture changed only later but even in the early 2000's there were InfoVis sessions at the VIS conference.
Removed a column of alternative unique identifiers that was unnecessary. The new column Q (previously column R) includes there references between IEEE VIS papers based on the IEEE article number.
We fixed a few DOIs and spelling of paper titles. Note that not all DOIs point to somewhere. The DOIs in our dataset are supposed to point to the IEEEXplore database. Yet, some are registered incorrectly and point to the IEEE Computer Society database instead. Some others are registered but point nowhere and some other papers have no DOI that we could find at all.
Many names in the deduped author column did not properly abbreviate the first name. This was fixed for most names - but following convention we left most Asian names as they were.We also now provide a worksheet that contains a one-column list of author names for your convenience. Check the tabs below the main data sheet or follow this link to get there:
Several author names did not have a space between the last name, and the first character of their first name. This was corrected for several papers.
The reference column used to include a mix of file names and IEEEXplore article numbers. It now only contains the article numbers. There were also a few instances of InfoVIs instead of InfoVis conference names that were cleaned up.
- Concerns the deduped author names and author names column
1-10 of 10